QUÉ ESTÁ SUCEDIENDO?

The common lactation curve pattern seen in dairy cattle is an increase in milk yield during early lactation, attaining a peak and then slowly decreasing until the end of lactation (Chang et al., 2001). This pattern, however, cannot be assumed to be uniform among other species. In fact, Cappio-Borlino et al. (1997) found significant effects of flock and feed supply combinations on lactation curves for dairy sheep. In addition to the effect of flock and feed supply, other factors which have significant effects on the pattern of the lactation curve include; lambing season (Peterson et al., 1990; Carta et al., 1995), lambing number (parity) (Portolano et al., 1997), breed (Hassan, 1995) and the length of time ewes have lambs at feet (suckling period) (Geenty, 1980).

There are two types of curves found in dairy sheep to explain the effect of stage of lactation on milk traits (Komprej et al., 2012). The first is common to intensive production systems and shows the daily milk yield increasing until a peak at three to five weeks of lactation. From the peak there is a gradual decline until the end of lactation. This persistency varies depending on factors such as the breed, environmental conditions and the individual.

The second type of lactation curve is considered atypical and is expressed in dairy ewes from extensive production systems. This type of curve indicates milk yield declining from the beginning of lactation right through to the end with no peak. A study by Oravcova et al. (2006) observed ewes of first-parity expressing a lower and earlier peak in the lactation curve, with a general better persistency than that of later parities. In addition, a study on the lactation curve of Valle del Belice dairy sheep in Slovakia showed significant effects of flock and feed supply combinations on test-day models of the lactation curve (Cappio-Borlino et al., 1997). With this said, both studies by Oravcova and Cappio-Borlino started milking at days 15 and 30 postpartum, which, according to the results from Peterson et al. (1994; 2005; 2006) was after the time of peak milk yield. Results from ewes milked daily for the first 12 days of lactation showed increasing milk yields (Peterson et al., 1994). Further studies by Peterson et al. (2005; 2006) cemented these findings, showing milk yields increasing from day one to seven. These results resemble the findings of McMillian et al. (2014a), where peak milk yield of dairy ewes grazing pasture was observed at the end of the first week postpartum. Conclusions from these studies indicate that milk yield peaks around day 10 or

12 and therefore, agree with the results of Oravcova and Cappio-Borlino, who presented declining milk yields from days 15 and 30.

Among the numerous factors previously mentioned, are additional environmental and genetic effects which strongly influence the shape of the lactation curve (Bauer et al., 2012). Environmental factors have been extensively documented in cattle, and more recently, in Slovakian sheep (Krupova et al., 2009). In order to distinguish between the genetic and environmental factors which influence daily milk yield, quantification of genetic and environmental variance parameters is required. These parameters, specific to each animal, can be treated as unobservable variables when undergoing genetic evaluation. When defining the breeding objective in an animal-breeding program, such parameters can be selected. Linear and non-linear mixed effect models obtain estimates of breeding values for modifying the shape of the lactation curve pattern by way of selection. Depending on the pattern of the curve, animals which show potential low performance can be culled.

The use of random regression methodology in this flock enabled the modelling of the lactation curve for each individual ewe with different lactation lengths. For example, the lactation curves for milk, fat and protein were obtained for a ewe which has 150 days in milk, as well as lactation curves for a ewe which has 210 days in milk (Figure 4.1). Random regression models are part of the routine genetic evaluation of most evaluation systems. This random methodology has several advantages (Jamrozik et al. 1997); a) permits the removal of environmental variation in phenotypic data on milk yield, since test-day milk yield considers the specific environmental effects for each herd-test record, b) grants a more accurate evaluation of animal with records, due to the use of a larger number of records per animal, and d) it facilitates the genetic evaluation of lactation persistency.

In the past, animal breeding strategies have been focused on maximising milk yield. However, this approach does not represent the best economic choice. A more economical choice seems to be improving farm profitability by extending the milking period, or, to change the components of milk. A way of doing this is to increase the lactation length or improve persistency of lactation. Lactation length and lactation persistency are both suitable ways of increasing total milk yield, however, a major drawback for implementing lactation persistency into a breeding goal is the difficulty of identifying an objective measure of such a trait (Cannas et al., 2002). There have been several attempts to measure persistency of

lactation (Swalve, 1995; Galal et al., 2008), which include a combination of measuring parameters and calculations to fit into lactation curves. Although these methods were suggested, none have been implemented as the standard model. A more recent, multivariate factor analysis has been used to estimate an index of lactation persistency in dairy cattle (Macciotta et al., 2002). Independent of the index used, heritability of lactation persistency is

low to moderate (0.10 – 0.30) (Chang et al., 2001). Care must be taken when including

lactation persistency in a breeding goal as its relationship with milk yield is variable. Some studies indicate that a lactation curve with a flatter pattern expresses a lower total milk yield, while others had the same production level (Macciotta et al., 2002).

Sheep milk has twice the amount of fat (7.62 vs. 3.67%) and protein (6.21 vs. 3.23%) than

cow’s milk (Jandal, 1996), and is mainly used for making cheese and yoghurt. As a result,

breeding programs should consider milk constituents rather than selecting solely on milk yield (Fuertes et al., 1998). Studies of lactation curves in dairy sheep demonstrate a relationship between milk yield and the levels of milk components (Carta et al., 1995; Oravcova et al., 2007). Selection for high milk yield is consequently accompanied by a decrease in milk fat and protein concentrations, while selection for high fat and protein content is known to result in lower milk yields (Oravcova et al., 2015). This relationship must be carefully examined when deciding on the most appropriate breeding objective as milk price in New Zealand may become dependent on milk yield and milk composition.

The mechanism for recording milk production data is known as a test-day approach. This model is characterised by the stage of lactation and days-in-milk. This test-day approach is achieved by a mathematical equation model. A fixed regression model is commonly used in milk-yield analysis (Hamann et al., 2004) as it is known to more accurately reflect the effect of days-in-milk on milk yield than that of a random regression model (Bauer et al., 2012). Most dairy sheep farms allow lambs to suckle during the early stage of lactation. This can lead to difficulties in accurately fitting complete lactation trajectories. Because of this, the lactation curve should only include milk yields and composition values from the milking period, and not include values for the entire lactation period.